East Asian Monsoon and paleoclimatic data analysis: a vegetation … · 2020. 6. 19. · and Mongolia from the last glacial maximum to Present and in particular the precipitation

Clim. Past, 4, 137–145, 2008www.clim-past.net/4/137/2008/© Author(s) 2008. This work is distributed underthe Creative Commons Attribution 3.0 License.

Climateof the Past

East Asian Monsoon and paleoclimatic data analysis:a vegetation point of view

J. Guiot1, Hai Bin Wu 2,3, Wen Ying Jiang4, and Yun Li Luo 5

1CEREGE, CNRS/Universite Paul Cezanne UMR 6635, BP 80, 13545 Aix-en-Provence cedex France2SKLL, Institute of Earth Environment, Chinese Academy of Sciences, Xian 710075, China3Institut des Sciences de l’Environnement, UQAM, Montreal PQ, H3C 3P8 Canada4Key Laboratory of Cenozoic Geology and Environment, Institute of Geology and Geophysics, Chinese Academy ofSciences, 100029 Beijing, China5Institute of Botany, Chinese Academy of Sciences, 100093, Beijing, China

Received: 10 December 2007 – Published in Clim. Past Discuss.: 20 February 2008Revised: 27 May 2008 – Accepted: 11 June 2008 – Published: 26 June 2008

Abstract. First we review several syntheses of paleodata(pollen, lake-levels) showing the climate variations in Chinaand Mongolia from the last glacial maximum to Present andin particular the precipitation increase at mid Holocene re-lated to enhanced monsoon. All these results concur to amuch enhanced monsoon on most of China during the firsthalf of the Holocene. Second we present, in some details, atemporal study of a core (Lake Bayanchagan, Inner Mongo-lia) located in an arid region at the edge of the present EastAsian Monsoon (EAM) influence and then sensitive to cli-matic change. This study involves pollen data together withother macro-remains and stable isotope curve to obtain a ro-bust climate reconstruction. This study shows a long wet pe-riod between 11 000 and 5000 years BP divided in two parts,a warmer one from 11 000 and 8000 (marked by large evap-otranspiration) and a cooler one more favourable to forestexpansion. Third, we present a spatial study based on pollendata only and covering all China and Mongolia at 6000 yearsBP, but using a mechanistic modelling approach, in an in-verse mode. It has the advantage to take into account en-vironmental context different from the present one (loweratmospheric CO2, different seasonality). This study showstemperature generally cooler than present one in southernChina, but a significant warming was found over Mongolia,and a slightly higher in northeast China. Precipitation wasgenerally higher than today in southern, northeast China, andnorthern Mongolia, but lower or similar to today in northwestChina and north China. Enhanced EAM was then found inthe southern half of China and in northeast China.

Correspondence to:J. Guiot([email protected])

1 Introduction

The past 21 000 years are a very interesting time period pe-riod as it contains two extreme states of the climate. TheLast Glacial maximum (LGM, 21 000 years BP) is a cold andgenerally dry period driven by enlarged ice sheets and low at-mospheric CO2. The mid-Holocene period (6000 years BP),generally warmer and wetter than the present one, is consid-ered as orbital forced period with perihelion in northern sum-mer/autumn and greater-than-present axial tilt (Berger, 1978)but free of major ice-sheet and relatively high CO2 (taking asreference the pre-industrial present time). These two peri-ods have been chosen as key time periods by the Palaeocli-mate Modelling Intercompraison Project, PMIP (Joussaumeand Taylor, 1995). The mid-Holocene, with its high summerinsolation, is a period of high land-sea contrast and conse-quently enhanced monsoon (Braconnot et al., 2002). It is ofparticular interest for climate modellers to test their simula-tions through palaeodata from the monsoonal regions.

The East Asian monsoon (EAM) is one of the most ac-tive components of the global climate system, influencing alarge area of China and its surrounding countries. In Chinaand surrounding countries, a megathermal period was recon-structed from 9500 to 4000 yr ago (Shi et al., 1993). How-ever, many recent studies have shown that Holocene climaticchanges were asynchronous across China (An, 2000; Anet al., 2006; He et al., 2004). The Holocene optimum was de-fined as EAM precipitation maximum, occurring ca. 11 000–9000 yr ago in northeastern China, 11 000–8000 yr ago innorth-central and northern east-central China, 8000–6000 yrago in central China, and ca. 3500 yr ago in southern China(An, 2000). The reason for debate on Holocene climaticvariations is that complexity of the EAM, and different re-sponses of environmental proxies to climatic changes (Weiand Gasse, 1999; Wang et al., 2003). Therefore, more

Published by Copernicus Publications on behalf of the European Geosciences Union.

http://creativecommons.org/licenses/by/3.0/

138 J. Guiot et al.: East Asian monsoon and paleovegetation

Fig. 1. Location of studied site and modern vegetation zones in China (afterJiang et al., 2006). I, Cold-temperate conifer forest; II,Temperate mixed conifer-broadleaved forest; III, Warm-temperate broadleaved deciduous forest; IV, Subtropical evergreen broadleavedforest; V, Tropical rainforest and seasonal rainforest; VI, Steppe; VII, Desert; VIII, Tibet-Qinghai cold and highland vegetation. The dashedand solid arrows indicate winter monsoon and the dominant direction of the summer monsoon precipitation belt, respectively.

precisely dated palaeo-records and improved quantitative re-construction are required to provide quantitative insights intothe processes of climatic changes, and their links to theEAM.

The goal of this paper is threefold. First we explore thetemporal variability of a record located in a sensitive re-gion at the northern edge of the EAM using a multiproxyapproach. Second we explore the spatial variability of theChinese climate at 6 ka BP, when EAM is assumed to be thestrongest. Third we illustrate a new methodology of climatereconstruction based on vegetation model inversion.

The temporal study is based on a core sampled in LakeBayanchagan (Inner Mongolia) (Jiang et al., 2006) (Fig. 1).This region is particularly sensitive to climate variations asit is located at the edge of the present EAM. Their resultssuggest that this region was dominated by steppe vegeta-tion throughout the Holocene, except for the period 9200to 6700 yr BP, when forest patches were relatively common.This period can then be correlated to enhanced EAM. Butthese findings need to be confirmed by a multiproxy analysis.We will synthesise in the first part of this paper an statisticalapproach based, in addition to pollen, on isotopic data andconcentration of a green algae species (Jiang et al., 20081).

1Jiang, W., Guiot, J., Wu, H., Chu, G., Yuan, B., Hatte, C., andGuo, Z.: Reconstruction of Holocene summer monsoon history us-ing d18O of carbonate, Pediastrum and pollen records from lake

This study will focus on the timing of this enhanced EAMperiod.

This approach based on detailed time series in a sensitiveregion will be completed by a spatial analysis based also onpollen data but done with the newest tools involving a pro-cess model able to relate vegetation and climatic variations(Luo et al., 20082). The strong feature of this approach isto be able to take into account the large differences existingbetween present and mid-Holocene conditions as (i) climateseasonality, possibly resulting in lack of modern analogues,or (ii) atmospheric CO2 close to pre-industrial concentrationbut significantly lower than the present one. This spatial anal-ysis will be first replaced in the context of previously pub-lished data syntheses at the sub-continental scale.

2 Data syntheses

An interesting story has been depicted byRen and Beug(2002) in the northern half of China (north of Yangtze river)for the whole Holocene. Forests generally expanded in

sediments in northern China, Global Planet. Change, submitted,2008.

2Luo, Y., Wu, H., Jiang, W., Guiot, J., and Sun, X.: Climaticchanges in China at the Last Glacial Maximum and mid-Holocene:reconstruction from pollen data using inverse vegetation modelling,in preparation, 2008.

Clim. Past, 4, 137–145, 2008 www.clim-past.net/4/137/2008/

J. Guiot et al.: East Asian monsoon and paleovegetation 139

the early Holocene times, reaching their maximum at 6 or4 ka BP, with a maximum in Central China, and then re-gressed during the late Holocene. An exception was foundfor northeast China where the maximum development of for-est occurred during the last 4000 or 2000 years. They con-cluded that, if the EAM enhancement seems to be responsi-ble of the forest expansion at the beginning of the Holocene,disturbance by human activities may be responsible of theforest decline after 6 ka BP.

This picture was completed by the study ofYu et al.(1998) who analysed the vegetation variations at the biomelevel for the whole China, but restricted at the 6 ka BP pe-riod. In eastern China at 6 ka BP, forest shifted northwards,with broadleaved evergreen forest extended about 300 kmand temperate deciduous forest about 500–600 km beyondtheir present northern limit. In northwestern China, the areaof desert and steppe vegetation was reduced as compared topresent. They concluded that these shifts were likely a re-sponse to enhanced Asian monsoon.

Lake levels data are less susceptible to be influenced byhuman disturbances.Yu et al.(2003) proposed a story of thelake levels since the Last Glacial Maximum (LGM, about21 ka BP). This compilation showed LGM conditions muchdrier than today in eastern China but somewhat wetter inwestern China. These east-west differential patterns of cli-mate conditions were completely different from the moderndry-wet conditions with a north-south opposition. Duringthe Holocene, at the mid-Holocene, both regions were wetterthan present. Modern dry conditions returned after 5–4 ka BPdepending on the region. Then if humans played a role inthe forest decline in the Late Holocene, they simply accentu-ated a climatic trend. Atmopheric general circulation models(AGCM) coupled with land surface process model showedthat the dry conditions in eastern China resulted from lesssummer precipitation due to the Pacific Subtropical High oc-cupying eastern China and the decline in the summer mon-soon.

More at north, in Mongolia,Tarasov et al.(1999) recon-structed, from pollen, warmer and wetter at 6 ka BP condi-tions for the northern part of the country, in agreement withhigher lake levels. In the central part of the country, warmerand drier conditions prevailed (inferred from pollen, no lakedata being available). But these dry conditions are likely dueto more evapotranspiration and not necessarily to less precip-itation.

3 A multi-proxy technique to reconstruct climate timevariability in Inner Mongolia

The syntheses presented above are based either on pollendata or lake lavels data. A multi-proxy approach is nowpresented to reinforce and precise these results (Jiang et al.,20081). It is based on a record taken from Lake Bayanchagan(115.21◦ E, 41.65◦ N, 1355 m a.s.l, Fig.1) in Inner Mongolia,

which is today almost completely dry due to anthropogenicwater use, with only small patches of shallow water main-tained by summer rain. It is situated at the current north-ern edge of the summer monsoon. The mean annual tem-perature in the area is about 3◦C, and total annual precipita-tion is 300–400 mm. About 70% of the precipitation occursduring the summer. The data used are pollen taxa countedfor 90 pollen assemblages and 2066 surface samples. Thetaxa are grouped into 17 plant functional types (e.g. borealevergreen conifers, steppics, grass, temperate summergreentrees, etc.) to reduce the number of variables and also toconsider together taxa which respond in the same way toclimatic variations. These plant functional types (PFT) areused to reconstruct climate by the modern analogue method(PFT-MAT) proposed byDavis et al.(2003) andJiang et al.(2006). The climatic variables considered are the tempera-ture of the coldest month (MTCO), the temperature of thewarmest month (MTWA), the annual precipitation (MAP),the ratio actual evapotranspiration over potential evapotran-spiration(α). These variables are calculated by linear inter-polation from meteorological stations (Jiang et al., 2006) andα is obtained from monthly temperature, precipitation andsunshine variables using the Priestley-Taylor equation (Pren-tice et al., 1992).

To these proxies, are added total pollen concentrations,Pediastrum(a green algae that indicate shallow lake water)concentrations andδ18O of authigenic carbonate, i.e. on the<40µm fraction (Jiang et al., 20081). These three proxiesshow a similar general pattern during the Holocene (Fig.2a).Before 11 000 cal yr BP, there is noPediastrumin the lake.Pollen concentrations are lower than 2×105 grains/ml. Allδ18O values of authigenic carbonate are between−3 and−1‰ VPDB. Similar values are found after 5 ka BP and inbetween, there is high concentrations of pollen andPedi-astrumand low δ18O values. As Jiang et al. (2008)1 haveshown that these three variables are controlled by balance be-tween precipitation and evaporation, they can be synthetizeda common signal, given here by their first principal compo-nent (Fig.2c).

Jiang et al. (2008)1 have used PFT-MAT constrained bythe first principal component PC1 (Fig.2c) as an indicator ofα, a variable directly related to the water stress. This con-strained analysis has already been proposed with differentproxies bySeret et al.(1992); Guiot et al.(1993); Cheddadiet al. (1996); Magny et al.(2001). For each fossil pollenspectrum, analogues were selected from the modern pollenspectra dataset subject to a broad consistency requirementaccording toα values. If we note the differenceδα betweenα of the analogue and the modernαo at the lake (56%), onlythe analogues i with aδαi compatible with PC1 at timet ,denotedCt , were retained. This compatibility is defined asfollows:

www.clim-past.net/4/137/2008/ Clim. Past, 4, 137–145, 2008


Fig. 2. Comparison of a few proxies ans the climatic reconstructions in Lake Bayanchagan (Inner Mongolia, China).(A) total pollen(105 grains ml−1) and Pediastrum concentration (104 grains ml−1), δ18O of authigenic carbonate multiplied by−1; (B) tree scores, i.e. sumof the square root of the arboreal taxa percentages;(C) first principal component of the three proxies of (A);(D) mean temperature of thecoldest month reconstruction;(E) mean temperature of the warmest month reconstruction;(F) total annual precipitation;(G) α, the ratiobetween actual and equilibrium evapotranspiration. The climate reconstructions are represented with the uncertainties, given by the range ofthe analogues. After Jiang et al. (2008)1.

Ct > 2 and δαi > 10%

Ct < −2 and δαi < −10%

−2 6 Ct 6 2 and −10%6 δαi 6 10% (1)

Figure 2 show the results obtained for the Lake Bayan-chagan core: the reconstructed climatic variables are com-pared with the constraint PC1 and the scores of the arbo-real pollen taxa (Fig.2b). This enables one to questionthe direct relationship often proposed between an increasein tree components of pollen assemblages and a warmerand wetter climate (Shi et al., 1993; Liu et al., 2002; Xiaoet al., 2004). So, the highest tree scores of trees during theHolocene in Lake Bayanchagan occurred between 8000 and5500 cal yr BP (Fig.2b). However, the peak period of treeswas not in phase with the warmest and wettest climate re-constructed between 11 000 and 8000 cal yr BP (Fig.2d–f),suggesting that a single climatic variable is not the trigger-ing factor. In contrast, variations in tree components andα

were consistent (Figs.2b and g).α is an integrated measureof annual amount of growth-limiting drought stress on plantsrelated to both temperature and precipitation, and is one ofprimary factors influencing vegetation distributions (Prenticeet al., 1992). The similarity in tree components andα varia-tions inferred from our study indicates that it is also the main

controlling factor for growth of trees over the Holocene in In-ner Mongolia.α does not reach its maximum before 8 ka BPeven if MAP is maximum because evaporation is too strong.The water stress is minimum only when temperature has de-creased by a few degrees. The most favourable period forforest development is then between 8 ka and 5 ka BP.

The MAP record during the Holocene at Lake Bayancha-gan is similar toδ18O records of stalagmite calcite fromDongge Cave and Shanbao Cave (Fig.1) in EAM regions(Dykoski et al., 2005; Shao et al., 2006). Shifts inδ18O val-ues of the stalagmite from the cave largely reflect changesin δ18O values of meteoric precipitation at the site, which inturn relates to changes in the amount of precipitation. Theδ18O results show that monsoon precipitation increased dra-matically at the start of the Holocene (∼11 500 cal yr BP) andremained high for∼6000 cal yr BP (Dykoski et al., 2005).This timing is consistent with other paleoclimatic recordsin EAM regions (Zhou et al., 2004, 2005). Both the LakeBayanchagan data and stalagmiteδ18O records from DonggeCave and Shanbao Cave show the termination of mon-soon precipitation maximum was abrupt between 6000 and4400 cal yr BP.

The second warm and humid period at Lake Bayanchagancentered at 6000 cal yr BP. This event was characterized byincreased MTCO, decreased MTWA and high precipitation



(Fig. 2d–f). These results agree with a marked increase inwinter temperatures across eastern China at 6000 cal yr BPestimated from pollen data (Yu et al., 1998) and simulated byclimatic model (Yu et al., 2003). The short-term cold eventbetween 8500 and 8300 cal yr BP was characterized by de-creases in both winter and summer temperature (Fig.2d–f).Even if such event has been recorded in several places andin particular in the GRIP and GISP2 records (Alley et al.,1997; Rohling and Palike, 2005), it cannot be considered assignificant in our reconstruction, as several such peaks arereconstructed during the Holocene.

4 A inverse modelling technique to reconstruct climatespatial variability in China

Multi-proxy approach is a way to produce robust paleocli-matic information but, as it is based on modern data using astatistical approach, it does not solve all the problems. Thereconstruction methods are built upon the assumption thatplant-climate interactions remain the same through time, andimplicitly assume that these interactions are independent ofchanges in atmospheric CO2. This assumption may lead toa considerable bias, as polar ice core records show that theatmospheric CO2 concentration has fluctuated significantlyover the past (EPICA, 2004). At the same time, a number ofphysiological and palaeoecological studies (Farquhar, 1997;Jolly and Haxeltine, 1997; Cowling and Sykes, 1999) haveshown that plant-climate interactions are sensitive to atmo-spheric CO2 concentration. Therefore, the use of mechanis-tic vegetation models has been proposed to deal with theseproblems (Guiot et al., 2000). Wu et al. (2007) have im-proved the approach based on the BIOME4 model to providebetter spatial and quantitative climate estimates from pollenrecords and correct for CO2 bias to pollen-based climate re-constructions in Eurasia and Africa. The same method isquickly presented here for Eastern Asian data.

4.1 Data and method

The pollen data used have been compiled by the BIOME6000project (Prentice and Jolly, 2000) for three key periods: 0 k,6 k and 21 ka BP to classify pollen assemblages into a set ofvegetation types. For the study described here, a subset con-taining 601 sample sites for 0 ka BP and 116 sites for 6 ka BPfrom China and Mongolia were used (MCPD, 2000, 2001;Tarasov et al., 1998). The selection of the 6 ka BP samples isbased following the BIOME6000 convention. Among them,84 sites have a good age control, i.e. either with at least twodates encompassing 6 ka BP at less than than 2000 years dis-tance.

BIOME4 is a physiological-process global vegetationmodel, with a photosynthesis scheme that simulates the re-sponse of plants to changed atmospheric CO2 and by ac-counting for the effects of CO2 on net assimilation, stomatal

conductance, leaf area index and ecosystem water balance. Itis driven by monthly temperature, precipitation, sunshine, byabsolute minimum temperature, CO2 concentration and soiltexture. The principle of the model inversion is to estimatethe input to BIOME4, the monthly climate, given that weknow some information related to the output of the model,biome scores derived from pollen in our case (Prentice et al.,1996). This inversion, which uses a Monte-Carlo-Markov-Chain algorithm to explore possible combinations of climateparameters, allows an assessment of the probability of dif-ferent anomalies, and therefore the investigation of differ-ent scenarios which may result in similar vegetation pattern.The procedure is described inWu et al. (2007). As Guiotet al.(2000), they showed that several solutions were possi-ble for the LGM climate in Western Europe where a mixtureof steppes and tundra existed. As these biomes have no clearanalogues today, reconstructions based on statistical methodswill tend to choose the least poor match or fail to find a realmatch. With the inverse modelling,Wu et al.(2007) showedthat a climate significantly warmer than inferred with mod-ern CO2 levels was the most probable. The overestimationof MTCO anomalies was about 10◦C. Moreover uncertain-ties were also underestimated with the statistical methods.

4.2 Validation

We present here an analysis of Chinese mid-Holocene data(Luo et al., 20082). In a first step, the ability of this inver-sion scheme to reproduce the modern climate of China isevaluated, using the 601 modern spectra available. The sta-tistical squared correlations (R2) between actual and recon-structed climate variables at the sample sites are presented inFig. 3. TheseR2 are very large, generally above 0.67, exceptfor MTWA which then does not appear to be a key factorto explain the modern vegetation distribution in China. Thestraight line between estimates and observations is expectedto have an intercept of 0 and a slope of 1. The slope is slightlybiased for MTWA, GDD and MAP. The intercepts are biasedfor MTCO, MTWA and MAT, showing a tendency to over-estimate the cold climates. There is also large error in esti-mating MAP andα in cold desert sites of the Tibet Plateau,whereα below 60% are frequently estimated below 20%, i.e.values typical of warmer deserts.

4.3 The 6000 yr BP climate

For the 6 ka BP period, the atmopsheric CO2 concentration isset to 270 ppmv (EPICA, 2004). The results (MAP, MAT,α)

are presented as maps of anomalies versus present climate(Fig. 4). Large circles indicate significant differences fromthe modern values. The results show that, in most of the sitesat 6 ka BP, the changes in precipitation andα were signifi-cantly different from modern values, while most of temper-ature changes are not. This is due to the larger uncertaintyon the reconstructed temperature, which indicates a larger



Fig. 3. Validation of the inverse BIOME4 modelling on the 601 samples of the modern database of China and Eurasia. The six reconstructedvariables are compared to the observed climate: MTCO (mean temperature of the coldest month reconstruction), present value =−20◦C,(MTWA) mean temperature of the warmest month reconstruction, present value = 17◦C, (GDD) growing degree days abve 5◦C, presentvalue = 1500◦ days, (α), the ratio between actual and equilibrium evapotranspiration, present value = 30%, (MAT) mean annual temperature,present value = 3◦C, (MAP) total annual precipitation, present value = 350 mm.

Fig. 4. Reconstruction of the climate in China 6000 years ago using inverse modelling method: (α), the ratio between actual and equilibriumevapotranspiration in %, (MAT) mean annual temperature◦C, (MAP) total annual precipitation in mm. All the values are given in departuresfrom present climate. Large circles indicate high significance levels (95%), small circles indicate no significance.



tolerance range of the vegetation to thermal variables whilehydrological variables were more limiting factors. Annualtemperature were generally lower than present one in south-ern China, but a significant warming was found over Mongo-lia, and a slight warming in northeast China.

Hydrological variables have a much more coherent distri-bution. MAP was generally higher than today in southern,northeast China, and northern Mongolia, but lower or similarto today in northwest China and north China.α was consid-erably higher than today in north China, and slightly higherthan present in northeast China. In contrast, drier conditionsare shown in northwest China and Mongolia.

Lake Bayanchagan is situated in a zone where most of thesites had a positive anomaly of MAP whereas a few ones hada negative one. This is broadly consistent with the recon-struction of Fig.2e where MAP was found 200 mm higherthan at present. The anomaly ofα for this zone is signif-icantly positive, between +15 and 30% in agreement withFig. 2f whereα was found 30% higher than at present. Forthese two variables, Lake Bayanchagan reconstruction pro-vide values at the upper limit of the inverse modelling. MATappears also higher than at present, in good agreement withthe reconstruction of Fig.2c–d. The reconstructions based onthe inverse modelling are then approximately consistent withthe Lake Bayanchagan, at least for the majority of surround-ing sites, but the multiproxy statistical approach infers valuesat the wetter limit of the inverse modelling. When comparedto Tarasov et al.(1999), Fig.4 shows also wetter and warmerconditions on northern Mongolia and warmer and drier con-ditions In the central part of the country.

5 Discussion

We have explored the temporal variability of the Lake Bayan-chagan record located in a sensitive region at the northernedge of the EAM. The use of a multiproxy approach coupledwith robust statistics have enlightened the complexity of theclimatic signal. A key problem in this respect is the timingof the monsoon enhancement. Monsoon increase is trans-lated in terms of increased precipitation. Then the period ofmaximum EAM occurred between 10.5 and 8 ka BP. A toorapid interpretation of the tree pollen curve should put thismaximum between 8 and 5 ka BP. It is clear that precipitationwas higher than at present time across the two periods. But,extension of forest depends as well of temperature than pre-cipitation, and our quantitative evaluation of several proxiesshow a more complex behaviour than Dongge and Shanbaocave series. This may also be due to spatial differences, thecaves being located at much more lower latitudes than thelake (Fig.1). This is confirmed by the post-5 ka BP decreasein the lake records where precipitation returns to the Late-Glacial level, while in the cave record,δ18O remains at anintermediate level. This might be explained by a rapid north-ward advance of the northern limit of the summer monsoon at

11.5 ka BP (beyond 41◦ N) followed by a slow retreat, fallingback south of Lake Bayanchangan by 5 ka, while the caves,being further south, remain under the monsoon influence.This illustrates well that Lake Bayanchagan, at the northernedge of the EAM zone, is a sensitive record of the monsoonsignal.

A second implication concerns the physical mechanisms.EAM enhancement is related to summer radiation which ismaximum at 9 ka BP and rapidly decreases to be at 6 ka BPon the same level than at 12 ka BP (Berger, 1978; An, 2000).When a large number of climate model simulations are com-pared (Braconnot et al., 2002), a robust feature is that theextension of the monsoon is related to the Eurasian conti-nent warming. This might explain why the 8–5 ka BP pe-riod is characterised by a slight decrease of EAM accom-panyed by a decrease of temperature more marked in thisnorthern lake than in lower latitudes. Maximum temperatureof the warmest month falls by 5◦C at 8 ka BP (but keep alevel above the present one), which shows a mitigation of theimpact on vegetation of the monsoon weakening by a sharpreduction of the evapotranspiration.

The analysis of the spatial variability of the Chinese cli-mate at 6 ka BP – even if the 6 ka period is not the periodof maximum monsoon enhancement – permits to replace thetiming found for Lake Bayanchagan in a larger context. Fig-ure 3 shows that some sites in the region of this lake havealready a reduced precipitation, whileα, which representsthe water availability for vegetation, is still higher than atpresent. This is still a period favourable to maintain forest,even with a precipitation reduction. Annual temperature dis-tribution shows higher values than at present time in northernChina, but lower in southern and central China where mon-soon had still a higher influence. This illustrates well the factthat northern China was more at that time under the influenceof the Eurasian continent while the rest of China was understill the influence of the ocean through the Pacific SubtropicalHigh.

A last point is the use of a new methodology of climatereconstruction based on vegetation model inversion. As al-ready mentionned, this mechanistic model offers the possi-bility to escape from too constraining modern conditions ashigh atmospheric CO2 concentrations or a climate seasonnal-ity different from modern one (in relation with insolation).The climatic maps obtained for 6 ka BP confirmed previousresults based on modern analogues, likely because CO2 con-centration is sufficiently high.Wu et al.(2007) have shownthat, for the Last Glacial Maximum conditions, biases are in-troduced by the fact that CO2 is sufficiently low to have lim-ited vegetation productivity in a comparable amplitude thanclimate change.



6 Conclusions

Various syntheses have been done on Chinese paleodata us-ing various methods. All converges to reconstruct intensifi-cation of EAM in China at 6 ka BP, especially on eastern partof China. Northern China cores indicate an intensified mon-soon between 10 and 5 ka BP. After 8 ka BP, a cooler climateinduced a less strong water stress, favouring the largest ex-tend of the forest. This two-step division of mid-Holocenehas been possible thanks to a multi-proxy approach enablingmore robust inference. Nevertheless, all approach involvingmodern analogues has its own limit when extrapolation isdone on periods with characteristics very different from thepresent reference period. Then the use of mechanistic mod-els in an inverse mode enables one to control the effect ofexternal variables, such as atmospheric CO2.

The results based on inverse modelling are coherent withthe previous syntheses. They show that a pattern of higherprecipitation is clear on eastern half of China. On westernpart of China, the situation is less contrasted with higher pre-cipitation on southwest and lower on northweast. The east-ern China situation is related to enhanced summer monsoonassociated with the Pacific Subtropical High bringing warmand most marine air from the West Pacific Ocean to east-ern China. The situation of southwestern China can be re-lated to the Indian summer monsoon bringing most marineair from the Indian Ocean to southern Tibetan Plateau andsouthern China lowlands. Northwestern regions are shelteredfrom these monsoon changes by the Tibetan Plateau and aredominated by the Westerlies and Asian winter monsoon. In-creased land-sea contrast due to higher summer insolationat mid-Holocene will then influence more strongly easternChina.

Acknowledgements.This research has been partly funded by agrant of the French Ministry of Research to two authors, the5th PCRD EU project MOTIF (EVK2-CT-2002-00153), by theEuropean Science Foundation, EUROCLIMATE/DECVEG andthe French ANR PICC.

Edited by: Ryuji Tada

References

Alley, R. B., Mayewski, P. A., Sowers, T., Stuiver, M., Taylor, K. C.,and Clark, P. U.: Holocene climatic instability: A prominent,widespread event 8200 yr ago, Geology, 25, 483–486, 1997.

An, C., Feng, Z.-D., and Barton, L.: Dry or humid? Mid-Holocenehumidity changes in arid and semi-arid China, Quaternary Sci.Rev., 25, 351–361, 2006.

An, Z. S.: The history and variability of the East Asian paleomon-soon climate, Quaternary Sci. Rev., 19, 171–187, 2000.

Berger, A.: Long-term variations of caloric insolation resultingfrom the earth’s orbital elements, Quaternary Res., 9, 139–167,1978.

Braconnot, P., Loutre, M. F., Dong, B., Joussaume, S., and Valdes,P.: How the simulated change in monsoon at 6 ka BP is relatedto the simulation of the modern climate: results from the Pale-oclimate Modeling Intercomparison Project, Clim. Dynam., 19,107–121, 2002.

Cheddadi, R., Yu, G., Guiot, J., Harrison, S. P., and Prentice, I. C.:The climate of Europe 6000 years ago, Clim. Dynam., 13, 1–9,1996.

Cowling, S. and Sykes, M.: Physiological significance of low atmo-spheric CO2 for plant-climate interactions, Quaternary Res., 52,237–242, 1999.

Davis, B. A. S., Brewer, S., Stevenson, A. C., Guiot, J., and con-tributors: The temperature of Europe during the Holocene recon-structed from pollen data, Quaternary Sci. Rev., 22, 1701–1716,2003.

Dykoski, C. A., Edwards, R. L., Cheng, H., Yuan, D. X., Cai, Y. J.,Zhang, M. L., Lin, Y. S., Qing, J. M., An, Z. S., and Revenaugh,J.: A high-resolution, absolute-dated Holocene and deglacialAsian monsoon record from Dongge Cave, China, Earth Planet.Sci. Lett., 233, 71–86, 2005.

EPICA: Eight Glacial cycles from an Antarctic ice core, Nature,429, 623–628, 2004.

Farquhar, G. D.: Carbon dioxide and vegetation, Science, 278,1411–1411, 1997.

Guiot, J., Beaulieu de, J. L., Cheddadi, R., David, F., Ponel, P.,and Reille, M.: The climate in Western Europe during the lastGlacial/Interglacial cycle derived from pollen and insect remains,Palaeogeogr. Palaeocl., 103, 73–93, 1993.

Guiot, J., Torre, F., Jolly, D., Peyron, O., Boreux, J. J., and Ched-dadi, R.: Inverse vegetation modeling by Monte Carlo samplingto reconstruct palaeoclimates under changed precipitation sea-sonality and CO2 conditions: application to glacial climate inMediterranean region, Ecol. Model., 127, 119–140, 2000.

He, Y., Theakstone, W., Zhang, Z., Zhang, D., Yao, T., Chen, T.,Shen, Y., and Pang, H.: Asynchronous Holocene climatic changeacross China, Quaternary Res., 61, 52–63, 2004.

Jiang, W., Guo, Z., Sun, X., Wu, H., Chu, G., Yuan, B., Hatte, C.,and Guiot, J.: Reconstruction of climate and vegetation changesof Lake Bayanchagan (Inner Mongolia): Holocene variability ofthe East Asian monsoon, Quaternary Res., 65, 411–420, 2006.

Jolly, D. and Haxeltine, A.: Effect of low glacial atmospheric CO2on tropical African montane vegetation, Science, 276, 786–788,1997.

Joussaume, S. and Taylor, K.: Status of the PMIP, Proceedingsof the First International AMIP Scientific Conference, WMO,Geneva, 425-430, 1995.

Liu, H. Y., Xu, L. H., and Cui, H. T.: Holocene history of deser-tification along the woodland-steppe border in northern China,Quaternary Res., 57, 259–270, 2002.

Magny, M., Guiot, J., and Schoellammer, P.: Quantitative recon-struction of Younger Dryas to mid-Holocene paleoclimates at LeLocle, Swiss Jura, using pollen and lake- level data, QuaternaryRes., 56, 170–180, 2001.

MCPD (Members of China Quaternary Pollen Database): Pollen-based Biome Reconstruction at Middle Holocene (6 ka BP) andLast Glacial Maximum (18 ka BP) in China, J. Integr. Plant Biol.,42, 1201–1209, 2000.

MCPD (Members of China Quaternary Pollen Database): Simula-tion of China Biome Reconstruction Based on Pollen Data from



Surface Sediment Samples. Journal of Integrative Plant Biology,J. Integr. Plant Biol., 43, 201–209, 2001.

Prentice, I. C. and Jolly, D.: Mid-Holocene and glacial-maximumvegetation geography of the northern continents and Africa, J.Biogeogr., 27(3), 507–519, 2000.

Prentice, I. C., Cramer, W., Harrison, S. P., Leemans, R., Monserud,R. A., and Solomon, A. M.: A global biome model based onplant physiology and dominance, soil properties and climate, J.Biogeogr., 19, 117–134, 1992.

Prentice, I. C., Guiot, J., Huntley, B., Jolly, D., and Cheddadi, R.:Reconstructing biomes from palaeoecological data: A generalmethod and its application to European pollen data at 0 and 6 ka,Clim. Dynam., 12, 185–194, 1996.

Ren, G. Y. and Beug, H. J.: Mapping Holocene pollen data andvegetation of China, Quaternary Sci. Rev., 21, 1395–1422, 2002.

Rohling, E. J. and Palike, H.: Centennial-scale climate cooling witha sudden cold event around 8,200 years ago, Nature, 434, 975–979, 2005.

Seret, G., Guiot, J., Wansard, G., Beaulieu, J. L. d., and Reille,M.: Tentative Palaeoclimatic Reconstruction Linking Pollen andSedimentology in La Grande Pile (Vosges, France), QuaternarySci. Rev., 11, 425–430, 1992.

Shao, X., Wang, Y., Cheng, H., Kong, X., Wu, J., and Lawrence, E.:Long-term trend and abrupt events of the Holocene Asian mon-soon inferred from a stalagmiteδ18O record from Shennongjiain Central China, Chinese Sci. Bull., 51, 221–228, 2006.

Shi, Y. F., Kong, Z. Z., Wang, S. M., Tang, L. Y., Wang, F. B., Yao,T. D., Zhao, X. T., Zhang, P. Y., and Shi, S. H.: Mid-HoloceneClimates and Environments in China, Global Planet. Change, 7,219–233, 1993.

Tarasov, P. E., Webb, T. I., Andreev, A. A., Afanas’ Eva, N. B.,Berezina, N. A., Bezusko, L. G., Blyakharchuk, T. A., Bo-likhovskaya, N. S., Cheddadi, R., Chernavskaya, M. M., Cher-nova, G. M., Dorofeyuk, N. I., Dirksen, V. G., Elina, G. A., Fil-imonova, L. V., Glebov, F. Z., Guiot, J., Gunova, V. S., Harrison,S. P., Jolly, D., Khomutova, V. I., Kvavadze, E. V., Osipova, I. M.,Panova, N. K., Prentice, I. C., Saarse, L., Sevastyanov, D. V.,Volkova, V. S., and Zernitskaya, V. P.: Present-day and mid-Holocene biomes reconstructed from pollen and plant macro-fossil data from the former Soviet Union and Mongolia: Globalchange – BIOME 6000, J. Biogeogr., 25, 1029–1053, 1998.

Tarasov, P. E., Guiot, J., Cheddadi, R., Andreev, a. A., Bezusko,L. G., Blyakharchuk, T. A., Dorofeyuk, N. I., Filimonova, L. V.,Volkova, V. S., and Zernitskaya, V. P.: Climate in NorthernEurasia 6000 Years Ago Reconstructed From Pollen Data, EarthPlanet. Sci. Lett., 171, 635–645, 1999.

Wang, B., Clemens, S., and Liu, P.: Contrasting the Indian andEast Asian monsoons: implications on geologic timescales, Mar.Geol., 201, 5–21, 2003.

Wei, K. and Gasse, F.: Oxygen Isotopes in Lacustrine Carbonates ofWest China Revisited: Implications for Post Glacial Changes inSummer Monsoon Circulation, Quaternary Sci. Rev., 18, 1315–1334, 1999.

Wu, H., Guiot, J., Brewer, S., and Guo, Z.: Climatic changesin Eurasia and Africa at the Last Glacial Maximum and mid-Holocene: reconstruction from pollen data using inverse vegeta-tion modelling, Clim. Dynam., 29, 211–229, 2007.

Xiao, J. L., Xu, Q. H., Nakamura, T., Yang, X. L., Liang, W. D., andInouchi, Y.: Holocene vegetation variation in the Daihai Lakeregion of north-central China: a direct indication of the Asianmonsoon climatic history, Quaternary Sci. Rev., 23, 1669–1679,2004.

Yu, G., Prentice, I. C., Harrison, S. P., and Sun, X. J.: Pollen-based biome reconstructions for China at 0 and 6000 years, J.Biogeogr., 25, 1055–1069, 1998.

Yu, G., Xue, B., Liu, J., and Chen, X.: LGM lake records fromChina and an analysis of climate dynamics using a modellingapproach, Global Planet. Change, 38, 223–256, 2003.

Zhou, W. J., Yu, X. F., Jull, A. J. T., Burr, G., Xiao, J. Y., Lu, X. F.,and Xian, F.: High-resolution evidence from southern China ofan early Holocene optimum and a mid-Holocene dry event dur-ing the past 18,000 years, Quaternary Res., 62, 39–48, 2004.

Zhou, W. J., Xie, S. C., Meyers, P. A., and Zheng, Y. H.: Re-construction of late glacial and Holocene climate evolution insouthern China from geolipids and pollen in the Dingnan peatsequence, Org. Geochem., 36, 1272–1284, 2005.


Documents

East Asian Monsoon and paleoclimatic data analysis: a vegetation … · 2020. 6. 19. · and Mongolia from the last glacial maximum to Present and in particular the precipitation